A Python Script to Merge Sanger Sequences

A Python script to merge Sanger sequences

Cen Chen1,2,3,*, Bingguo Lu4,*, Xiaofang Huang5, Chuyun Bi5, Lili Zhao6, Yunzhuo Hu5, Xuanyang Chen5, Shiqiang Lin5 and Kai Huang1 1 Clinical Center for Human Genomic Research, Union Hospital, Huazhong University of Science and Technology, Wuhan, China 2 Department of Radiology, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China 3 Department of Cardiovascular Diseases, Union Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China 4 Provincial University Key Laboratory of Cellular Stress Response and Metabolic Regulation, College of Life Science, Fujian Normal University, Fuzhou, China 5 Key Laboratory of Crop Biotechnology, Fujian Agriculture and Forestry University, Fujian Province Universities, Fuzhou, China 6 State Key Laboratory of Infectious Disease Prevention and Control, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control, Beijing, China * These authors contributed equally to this work.

ABSTRACT Merging Sanger sequences is frequently needed during the gene cloning process. In this study, we provide a Python script that is able to assemble multiple overlapping Sanger sequences. The script utilizes the overlapping regions within the tandem Sanger sequences to merge the Sanger sequences. The results demonstrate that the script can produce the merged sequence from the input Sanger sequences in a single run. The script offers a simple and free method for merging Sanger sequences and is useful for gene cloning.

Subjects Bioinformatics, Genetics, Genomics, Molecular Biology, Computational Science Keywords Sanger sequencing, Merge, Python script Submitted 15 December 2020 Accepted 5 April 2021 Published 27 April 2021 INTRODUCTION Corresponding authors Gene cloning is customarily required to study the gene functions in vivo and in vitro. Shiqiang Lin, For the most part, the gene of concern is ligated to a vector with a canonical method of [email protected] restriction enzyme cutting and ligation, or the new-fashioned technique of seamless Kai Huang, [email protected] ligation (Okegawa & Motohashi, 2015). To ensure that the gene sequence is correct within Academic editor Joseph Gillespie the constructed plasmid, the Sanger sequencing is utilized to sequence the gene, and the Additional Information and results are aligned and checked (Zimmermann et al., 1988). Declarations can be found on Sanger sequencing is a DNA sequencing technology by incorporating chain-terminating page 8 dideoxynucleotides, which are ﬂuorescently labeled and can be read-out by automated DOI 10.7717/peerj.11354 sequencing machinery. Sanger sequencing has been widely applied to determine the DNA Copyright sequence since it was reported ﬁrst time by Sanger et al. (1977) and Sanger, Nicklen & 2021 Chen et al. Coulson (1977). Though the next-generation sequencing has started a new era, Sanger Distributed under sequencing is still one of the most popular methods due to its reliability, affordability, Creative Commons CC-BY 4.0 and feasibility (Schuster, 2008). However, the number of nucleotides determined by one

How to cite this article Chen C, Lu B, Huang X, Bi C, Zhao L, Hu Y, Chen X, Lin S, Huang K. 2021. A Python script to merge Sanger sequences. PeerJ 9:e11354 DOI 10.7717/peerj.11354 Sanger sequencing reaction is around 1,000. For those genes with lengths of only several hundred nucleotides, it is possible to go through each one of the whole sequences with one single Sanger sequencing reaction, and the results can be directly aligned to the correct gene files. However, in other cases, the lengths of the genes exceed one thousand nucleotides, beyond the scope of one single Sanger sequencing reaction. To get the full-length sequence of a target gene, it is necessary to carry out DNA walking sequencing using a new primer based on the previous sequencing result. To sequence a large gene, several reactions might be required in both forward and reverse directions. These results have to be aligned to the correct gene file for confirmation. It is preferred to merge all the walking results before alignment with the correct gene file, rather than aligning each walking result with the target gene file separately (Tang et al., 2020a, 2020b), especially for those large genes, which may be over 10,000 bps in length. To merge the multiple walking results, several commercial software can be applied, such as DNASTAR (DNASTAR, Inc., Madison, WI, USA), DNAMAN (Lynnon Biosoft, San Ramon, CA, USA), Vector NTI (Thermo Fisher Scientific Inc., Denver, CO, USA), and SnapGene (GSL Biotech LLC., San Diego, CA, USA), which are robust, however, expensive. For protein science and molecular biology, sequencing is mostly used to confirm whether the subcloning or mutagenesis is correct. Thereby, some functions of these commercial software are not essential, and the price of these software might be too high for the routine work of molecular cloning. In terms of free software, there is an online tool that is able to merge overlapping long sequence fragments based on the program merger in the “EMBOSS” suite (Bell & Kramvis, 2013; Rice, Longden & Bleasby, 2000). Nevertheless, the web tool relies on web access and server status. “Cap3” is a free and effective program for merging DNA fragments, however, the command-line is complicated for users who are not familiar with commands of “Cap3” (Huang & Madan, 1999). “Staden”, containing a graphic user interface, provides a free and powerful package for analyzing giant genome sequencing by combining the “SPIN” and “EMBOSS” (Staden, Judge & Bonfield, 2003), again, this package is complicated for some users. Since the first trial of the assembly of DNA sequence by computer, the algorithms have developed rapidly, which can be categorized into three classes: the first is overlap-layout- consensus method (Batzoglou, 2005), which applies an overlap graph; the second is de Bruijn graph roads (Idury & Waterman, 1995; Pevzner, 1989; Pevzner, Tang & Waterman, 2001), which uses a k-mer graph; and the last is greedy algorithm, which uses a k-mer or an overlap graph (Pop, 2009; Pop & Salzberg, 2008). Most of these algorithms are coded by C, which is challenging for biologists with less programming backgrounds. Here, we introduce a Python script capable of merging multiple overlapping Sanger sequencing files by employing the alignment module of Biopython (Cock et al., 2009). To visualize the critical parameters of the merging process, we print the output of alignment and show the critical junction within the merged sequence. The script is useful for merging the Sanger sequences and increasing the efficiency of subcloning.

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 2/10 METHODS Running environments and input files To run the script “Merge_Sanger_v2.py”, the macOS Catalina 10.15.3 or higher (Apple Inc., Cupertino, CA, USA) is needed. Python 3.7.3 (https://www.python.org/), Biopython 1.7.4, and “EMBOSS” 6.6.0 are necessary for running the script in this study (Cock et al., 2009; Rice, Longden & Bleasby, 2000). The users need to install the above three packages before running the script. Here are the sequencing result files of the pET28a- SV40NLS-hCas9-SV40NLS, which is originated from PX458 (Ran et al., 2013) (Supplemental Information S1)(Fig. 1). All input files are in the seq format. Of note, the first four characters of the input file name must consist of three numbers (000, 001, 002, 003, …) plus one capital letter F or R (F for forward Sanger sequencing and R for reverse Sanger sequencing). Adding extra four characters to the original filename provided by the sequencing company is suggested. The numbers in the filenames should be ordered from small to big along with the 5′ to 3′ of the sense strand.

Running script Please follow the steps below to run the script.

1. Make a new directory and copy the sequencing result ﬁles of the above section (Running environments and input ﬁles) to the newly made directory. Also, copy the script “Merge_Sanger_v2.py” (Supplemental Information S2) to the directory. 2. Open a new terminal and “cd” to the directory. Run the following command. Notice that a space between every two items is needed and the backward slash is used for multiline command input.

python3.7 Merge_Sanger_v2.py 001F_sequence_to_be_merged.seq \ 002F_sequence_to_be_merged.seq 003F_sequence_to_be_merged.seq \ 004R_sequence_to_be_merged.seq 005R_sequence_to_be_merged.seq \ 006R_sequence_to_be_merged.seq

The Sanger sequencing result files must be input from lower to higher according to the first three numbers of the file name. It is convenient to input the first three numbers and then use the tab key to fill the rest of the file name automatically. Our script is able to merge dozens of (up to 1,000) Sanger sequencing result files at a time, which is far enough for usual needs in labs.

3. After running ﬁnishes, a folder named “merged_sequence”, which contains the merged sequence ﬁle with the name “merged.seq”, appears.

Manual validation of the merged Sanger sequences First, all the Sanger sequence ﬁles are renamed and changed to fasta format with TextEdit. Those Sanger sequence ﬁles in the reverse sequencing direction are transformed to the reverse complement sequences with “EMBOSS” command “revseq”. All the Sanger

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 3/10 Figure 1 The test ﬁles used for running the script. Full-size  DOI: 10.7717/peerj.11354/ﬁg-1

sequence files are now in the forward sequencing direction in fasta format. Then, the “EMBOSS”“needle” is utilized to align each pair of tandem sequence files, and the command “less” is used to open the result files of alignments. The final merged sequence file is generated by combining the Sanger sequences based on the overlapping regions. The manually combined sequence file is used to align with the script’s merged sequence, and the time consumed by the manual method is compared with that used by the script.

Comparison with CAP3 and Fragment Merger for merging Sanger sequences “CAP3” is a free and powerful software for contig assembly during shotgun sequencing of genome projects (Huang & Madan, 1999). It is also incorporated in commercial software to assemble Sanger sequencing results, for example, SnapGene (GSL Biotech LLC., San Diego, CA, USA). For comparison, the on-line version of the “CAP3” program at the website (http://doua.prabi.fr/software/cap3) is used. With regard to the example input files, we first try to assemble only two tandem Sanger sequences, then three tandem Sanger sequences, and so on. We also compared “Fragment Merger” (Bell & Kramvis, 2013)at the website (http://hvdr.bioinf.wits.ac.za/fmt/) using the example Sanger files. The example files in seq format were changed to FASTA format and then running of the “Fragment Merger” is carried out following the instructions on the webpage using default parameters. Similarly, we start from assembly of only two tandem Sanger sequences, then three tandem Sanger sequences, and so on.

Algorithm Our script heavily relies on excellent pre-existing packages, i.e., “EMBOSS” (Rice, Longden & Bleasby, 2000) and Biopython (Cock et al., 2009), which are versatile and extremely useful to molecular biologists world around. The script begins to transform the input Sanger sequence files, which might be in forward or reverse sequencing directions, into the new Sanger sequence files in forward sequencing direction. Our algorithm takes advantage of the overlapping regions within the tandem Sanger sequences, which are commonly no less than 200 bps (Fig. 2). The alignment module of the Biopython is called to conduct the needle alignment of the two consecutive Sanger sequence files. The script then parses the output file of needle alignment to seek for the first 100% identity alignment fragment that is no less than 50 bps and records the critical fragments of the

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 4/10 000F_sequence

001F_sequence

002F_sequence

00nF_sequence

Figure 2 A diagrammatic view of merging multiple Sanger sequences. The thick blue lines show the multiple (no less than two) Sanger sequences to be merged, with the color depth indicating the sequencing quality, which is high in the middle part and low in both ends. The thick orange lines represent the first line of full consensus (50 bps long) appearing in the needle output file for the two consecutive Sanger sequences. The thin blue arrow above the thick blue line indicates the part to be extracted for joining to the full-length sequence, which is shown by the long thin blue arrow at the bottom. All the Sanger sequences are preprocessed to 5′–3′ direction, aligned and merged, thus the direction of all thin arrows is 5′–3′. Full-size  DOI: 10.7717/peerj.11354/fig-2

Figure 3 The directory showing the output ﬁle after merging the Sanger sequences. Full-size  DOI: 10.7717/peerj.11354/ﬁg-3

corresponding Sanger sequences. These critical fragments are then used to extract the bases from each Sanger sequence for merging to the full-length sequence.

RESULTS When the script finishes one run, the merged Sanger sequence “merged.seq” is stored in a new folder named “merged_sequence” (Fig. 3). Besides, there are outputs on the screen showing the needle alignment and the critical junction of the consensus region between the two adjacent Sanger sequences (Fig. 4), which clarifies the working process of merging. Further, the merged sequence is shown on the screen. The results showed that the running process of the script is fast, and the output is clear. Then, we conducted a manual operation of merging the Sanger sequences (Supplemental Information S3). The manual method is labor-intensive and tedious. It took about one hour to carry out the manual merging of the seven Sanger sequencing files, whereas the script accomplishes the same job within seconds after command input. Theoretically, our script works in the same way as the manual operation, showing that our script conforms to molecular biologists’ intuition and experience when merging the Sanger sequences. Actually, our script prints the working process including the results of each

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 5/10 Figure 4 The screenshots of the script run showing the partial output of the needle alignment (A) and the ﬁrst consensus line and the merged DNA sequence (B). Full-size  DOI: 10.7717/peerj.11354/ﬁg-4

“EMBOSS”“needle” alignment and the boundary used for fragment joining so that the user knows what’s going on during the Sanger sequence merging (Supplemental Information S4). Moreover, the script works as accurately as the manual operation and saves time in the Sanger sequence merging. Besides, the result generated manually (Supplemental Information S3) is the same as that obtained using the script. We also compared the script with “CAP3”, a free, well-known, and widely-used software for contig assembly. First, using the test input ﬁles, we tried the script to assemble two, three, four, and up to seven tandemly arranged Sanger sequences. As was expected, the script worked well in all the above situations. However, when we tried merging the same example Sanger sequences with the online version of “CAP3” (Huang & Madan, 1999), out of expectation, “CAP3” could not merge several two tandem Sanger sequences, e.g., 000F_sequence_to_be_merged.seq and 001F_sequence_to_be_merged.seq could not be merged by “CAP3”. Nevertheless, the quality of 000F_sequence_to_be_merged.seq and 001F_sequence_to_be_merged.seq fulﬁlled the requirement for Sanger sequences merging. It seems that “CAP3” is not completely well suited to merging the example Sanger sequences in its default parameters. When using the online “Fragment Merger” to assemble the example Sanger sequences (Supplemental Information S5), we found that the assembled sequence was not the same as that obtained by manual operation. This is probably because the fact that “Fragment

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 6/10 Free of charge

Source code Based on and annotation Biopython and are provided EMBOSS needle Easily adjusted based on Reliable personal Advantages performance requirements

Mimicking the High speed assembly logic Time saving of Sanger Visualizing sequences process

Figure 5 Advantages of the script. Full-size  DOI: 10.7717/peerj.11354/ﬁg-5

Merger” uses the “EMBOSS”“merger”, in which the issue of the disputed bases needs to be solved by reading the corresponding peaks in the chromatogram ﬁles. In comparison, our script can deal with all tested situations since the script adopts the same logic as the manual operation when merging the Sanger sequences, which agrees with the inherent characteristics of Sanger sequencing (sequencing quality from low to high to low in a reaction) and the reality of experimental operations. Notably, the commercial solutions, such as DNASTAR, DNAMAN, Vector NTI, and SnapGene, cost hundreds to thousands of dollars’ license fees per year, which is expensive for many small groups. In contrast, our script is open source and free of charge, and fulﬁlls the basic requirements of molecular cloning (Bi et al., 2020). Moreover, compared to the black box of commercial software, we provide the source code, written in Python language that resembles natural languages and annotated in detail, which is relatively easy to learn and allows the users to tailor the script according to their needs. Together, the advantages of our script are free of charge, open-source, reliable, and high speed (Fig. 5).

DISCUSSION Merging Sanger sequences is common and necessary work during gene cloning. The “EMBOSS” command “merger” is capable of generating a merged sequenced based on the global alignment of Needleman and Wunsch (Rice, Longden & Bleasby, 2000). To merge two overlapping Sanger sequences, the global alignment might bring in mismatches due to the decrease of signal-to-noise ratio during the later stage of the Sanger

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 7/10 sequencing reaction and gel running. According to the local sequence quality score, the “EMBOSS”“merger” deals with the disputed bases, which is rather complicated and often brings in the unwanted bases. To address this issue, we look for the first full consensus line within the output of needle alignment and pick up the fragments to join the merged sequence according to the variations of the signal-to-noise ratio of the Sanger sequencing reaction, which avoids dealing with the dispute settlement for the mismatch bases. Besides, using “EMBOSS” “needle”, our algorithm can align tandemly arranged Sanger sequences and calculate the boundary of sequences to merge, potentially minimizing the influence of repeated sequences utmostly. Interestingly, while we are trying to simplify the assembly of Sanger sequences using Python, Chao et al. reported an automated R package, “sangeranalyseR”, for analyzing the Sanger sequencing data (Chao et al., 2020; Hill et al., 2014). The “sangeranalyseR” is an interactive program based on the R/bioconductor that focuses on analyzing Sanger sequence data in ABI format (Gentleman et al., 2004). Together with the program reported in this manuscript, it reflects that the researchers require various programs based on different coding languages for merging DNA fragments since each researcher has specific background and requirements.

CONCLUSIONS In sum, we provide an efficient and straightforward method to merge multiple Sanger sequences using Python, which can be run at a local computer and fulfills the basic requirements of subcloning and protein science. Our script mimics the manual assembly of the Sanger sequences like a molecular biologist, which is in line with the researchers’ experience and intuition in the fields of gene and protein.

ADDITIONAL INFORMATION AND DECLARATIONS

Funding This study is supported by the National Natural Science Foundation of China, Grant/ Award Number: 91949201 (Kai Huang). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Grant Disclosures The following grant information was disclosed by the authors: National Natural Science Foundation of China: 91949201.

Competing Interests The authors declare that they have no competing interests.

Author Contributions Cen Chen conceived and designed the experiments, performed the experiments, analyzed the data, prepared ﬁgures and/or tables, and approved the ﬁnal draft.

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 8/10 Bingguo Lu conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, and approved the final draft. Xiaofang Huang performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Chuyun Bi performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Lili Zhao performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Yunzhuo Hu performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Xuanyang Chen performed the experiments, authored or reviewed drafts of the paper, and approved the final draft. Shiqiang Lin conceived and designed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft. Kai Huang conceived and designed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the paper, and approved the final draft.

Data Availability The following information was supplied regarding data availability: The example input files, the Python script, the readme file, the manual validation of the merging of the example input files, the screen output of the script run, and the comparison with Fragment Merger for Sanger sequence merging are available in the Supplemental Files.

Supplemental Information Supplemental information for this article can be found online at http://dx.doi.org/10.7717/ peerj.11354#supplemental-information.

REFERENCES Batzoglou S. 2005. Algorithmic challenges in mammalian whole-genome assembly. In: Jorde LB, Little PFR, Dunn MJ, Subramaniam S, eds. Encyclopedia of Genetics, Genomics, Proteomics and Bioinformatics. Hoboken: John Wiley & Sons. Bell TG, Kramvis A. 2013. Fragment merger: an online tool to merge overlapping long sequence fragments. Viruses 5(3):824–833 DOI 10.3390/v5030824. Bi C, Huang X, Tang D, Shi Y, Zhou L, Hu Y, Chen X, Qi S, Lin S. 2020. A python script to design site-directed mutagenesis primers. Protein Science 29(4):1054–1059 DOI 10.1002/pro.3806. Chao K-H, Barton K, Palmer S, Lanfear R. 2020. sangeranalyseR: simple and interactive analysis of Sanger sequencing data in R. bioRxiv DOI 10.1101/2020.05.18.102459. Cock PJ, Antao T, Chang JT, Chapman BA, Cox CJ, Dalke A, Friedberg I, Hamelryck T, Kauff F, Wilczynski B, de Hoon MJ. 2009. Biopython: freely available Python tools for computational molecular biology and bioinformatics. Bioinformatics 25(11):1422–1423 DOI 10.1093/bioinformatics/btp163. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, Hornik K, Hothorn T, Huber W, Iacus S, Irizarry R, Leisch F, Li C, Maechler M, Rossini AJ, Sawitzki G, Smith C, Smyth G, Tierney L, Yang JYH, Zhang J. 2004.

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 9/10 Bioconductor: open software development for computational biology and bioinformatics. Genome Biology 5(10):R80 DOI 10.1186/gb-2004-5-10-r80. Hill JT, Demarest BL, Bisgrove BW, Su YC, Smith M, Yost HJ. 2014. Poly peak parser: method and software for identification of unknown indels using sanger sequencing of polymerase chain reaction products. Developmental Dynamics 243(12):1632–1636 DOI 10.1002/dvdy.24183. Huang X, Madan A. 1999. CAP3: a DNA sequence assembly program. Genome Research 9(9):868–877 DOI 10.1101/gr.9.9.868. Idury RM, Waterman MS. 1995. A new algorithm for DNA sequence assembly. Journal of Computational Biology 2(2):291–306 DOI 10.1089/cmb.1995.2.291. Okegawa Y, Motohashi K. 2015. A simple and ultra-low cost homemade seamless ligation cloning extract (SLiCE) as an alternative to a commercially available seamless DNA cloning kit. Biochemistry Biophysics Reports 4:148–151 DOI 10.1016/j.bbrep.2015.09.005. Pevzner PA. 1989. 1-Tuple DNA sequencing: computer analysis. Journal of Biomolecular Structure and Dynamics 7(1):63–73 DOI 10.1080/07391102.1989.10507752. Pevzner PA, Tang H, Waterman MS. 2001. An Eulerian path approach to DNA fragment assembly. Proceedings of The National Academy of Sciences of The United States of America 98(17):9748–9753 DOI 10.1073/pnas.171285098. Pop M. 2009. Genome assembly reborn: recent computational challenges. Briefings in Bioinformatics 10(4):354–366 DOI 10.1093/bib/bbp026. Pop M, Salzberg SL. 2008. Bioinformatics challenges of new sequencing technology. Trends in Genetics 24(3):142–149 DOI 10.1016/j.tig.2007.12.006. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F. 2013. Genome engineering using the CRISPR-Cas9 system. Nature Protocols 8(11):2281–2308 DOI 10.1038/nprot.2013.143. Rice P, Longden I, Bleasby A. 2000. EMBOSS: the European molecular biology open software suite. Trends in Genetics 16(6):276–277 DOI 10.1016/S0168-9525(00)02024-2. Sanger F, Air GM, Barrell BG, Brown NL, Coulson AR, Fiddes CA, Hutchison CA, Slocombe PM, Smith M. 1977. Nucleotide sequence of bacteriophage phi X174 DNA. Nature 265(5596):687–695 DOI 10.1038/265687a0. Sanger F, Nicklen S, Coulson AR. 1977. DNA sequencing with chain-terminating inhibitors. Proceedings of The National Academy of Sciences of The United States of America 74(12):5463–5467 DOI 10.1073/pnas.74.12.5463. Schuster SC. 2008. Next-generation sequencing transforms today’s biology. Nature Methods 5(1):16–18 DOI 10.1038/nmeth1156. Staden R, Judge D, Bonfield J. 2003. Analyzing sequences using the staden package and EMBOSS. In: Krawetz SA, Womble DD, eds. Introduction to Bioinformatics. Totowa, NJ: Humana Press, 393–410. Tang D, Chen C, Liao G, Liu J, Liao B, Huang Q, Chen Q, Zhao J, Jiang H, Duan J, Huang J, Wang K, Wang J, Zhou C, Chu W, Li W, Sun B, Li Z, Dai L, Fu X, Cheng W, Xue Y, Qi S. 2020a. Structural and functional insight into the effect of AFF4 dimerization on activation of HIV-1 proviral transcription. Cell Discovery 6:7 DOI 10.1038/s41421-020-0142-6. Tang D, Sheng J, Xu L, Zhan X, Liu J, Jiang H, Shu X, Liu X, Zhang T, Jiang L, Zhou C, Li W, Cheng W, Li Z, Wang K, Lu K, Yan C, Qi S. 2020b. Cryo-EM structure of C9ORF72-SMCR8- WDR41 reveals the role as a GAP for Rab8a and Rab11a. Proceedings of The National Academy of Sciences of The United States of America 117(18):9876–9883 DOI 10.1073/pnas.2002110117. Zimmermann J, Voss H, Schwager C, Stegemann J, Ansorge W. 1988. Automated Sanger dideoxy sequencing reaction protocol. FEBS Letters 233(2):432–436 DOI 10.1016/0014-5793(88)80477-0.

Chen et al. (2021), PeerJ, DOI 10.7717/peerj.11354 10/10